The purpose of this study is to investigate the influence of low dose oral contraceptives on bone resorption in perimenopausal women as determined by urinary markers of bone metabolism. Twenty healthy perimenopausal women will be enrolled. The perimenopausal women will meet all inclusion criteria prior to enrollment. Inclusion criteria Will include 1) age between 45-50 years, 2) change in menstrual cycle interval, 3) complaints of hot flashes, 4) elevated follicular or luteal phase FSH and 5) serum creatinine < 1.5 mg/dI. Patients who have been longer than 6 months without a menstrual cycle or have taken hormonal therapy within the preceding 6 months will be excluded. In addition, women who smoke, have a history of taking bisphosphonates, calcitonin, or calcitriol, or who have a history of parathyroid disease will be excluded. Two baseline urine specimens from the first morning void will be collected on two separate days from patients enrolled in the study. Urinary concentrations of deoxypyridinoline crosslinks will be measured using Pyrlinks-D (Metra Biosystems, Mountain View, CA) enzyme-linked immunoassay. Urinary concentrations of Ntelopeptides will be measures using Osteomark (Ostex, Seattle, WA) enzyme-linked immunoassay. Concentrations of deoxypyridinoline crosslinks and N-telopeptides from the two baseline urine specimens will be averaged to determine the baseline excretion prior to initiating hormonal therapy. Patients will then be given Loestrin (Parke-Davis, Morris Plains, NJ), containing 20 mcg ethinyl estradiol with I mg norethindrone acetate the following Sunday. Urinary specimens obtained from the first morning void will be collected on a weekly basis for a total of four weeks after initiation of Loestrin. The urinary specimens will be assayed for both deoxypyridinoline crosslinks and Ntelopeptides. Concentrations of urinary deoxypyridinoline crosslinks and N-telopeptides prior to and after each successive week of oral contraceptive pill use win then be compared. A one-tailed Student's t-test will be used for statistical analysis. Significance will be defined as a P value < 0.05. The power of the study is anticipated to be greater than 0.80 based on previous responses to estrogen therapy on urinary niarkers of bone metabolism.